Northern ELectric N293 vs N393: which one did I just buy?

[DuinoSoar]

I just received the phone about 2 hours ago.  It looks in pretty good shape except for the switch hook (which I already knew about).  The switch contact leaf springs are quite badly bent out of shape and I am not sure yet whether I can bend them back into operating condition or not.  It looks like somebody tried to fix it at some time after the seller's pictures were posted, because the leaf springs are bent differently than shown in the picture (see the pics in my opening contribution to this thread).

Also, as we guessed, the picture of the "innards" were a bit misleading about the wires connected to the switch.  It turns out that the bottom and top contacts are NOT bridged by the white/red wire.  The white/red wire is connected only to the bottom solder lug, and the black wire is soldered to the top one; nothing is connected to the middle solder lug.

Strangely, the receiver element looks like a transmitter element (microphone) from a 500 set which does not make sense to me, because vintage 500 set microphones are carbon, not dynamic.  The date code on the receiver element is 3-57 and the black metal front has "HA" embossed into it. It may be dynamic; I don't know for sure yet.  In any case, I will connect the phone and try it out over the weekend, to ensure the transmitter, receiver, dial and ringer are all in good working order, and then I will try to see if I can fix the spring contacts on the switch hook. If not, I will contact Jeff at OPW to see if I can get a reproduction switch.

Once I get it all working, I am debating whether to mount it on a wall in my apartment, or to add some nice, thick soft-plastic feet to the bottom to make it into a "desktop" phone.  (I am leaning toward wall-mounting it though.)

Later on, I will add a proto-board circuit for the pulse-to-DTMF converter.

BTW, my Rogers Ignite gateway (which includes the "home-phone" service port) accepts decadic pulse dialling, but only for originating a call; it will not translate pulses to DTMF for responding to interactive voice menus.

Best regards,
  Ed ("The DuinoSoar")

[HarrySmith]

How about some pictures? We LOVE pictures

[DuinoSoar]

How about some pictures? We LOVE pictures

The phone looks just like the pictures from the seller, which I posted in my original message when I opened this thread.  (Well, except that the switch springs are bent a little differently now than what was shown in the seller's pics.)  I will probably take some pics over the weekend as I work on trying to fix the switch.

Is there anything in particular you would like me to focus on or highlight in pictures?

[HarrySmith]

Nothing in particular. Maybe just document your progress?

[TelePlay]

IIRC, you can take the switch hook pileup apart and work on each spring by themselves.

Just remember the order in which it came apart.

[poplar1]

Here is a WE 293A that shows the 3-leaf switchhook:

[DuinoSoar]

It has been almost a half year since the last post on this thread.  Just before that last post from @poplar1, on Sept 28, "real life" interrupted, and I could not do anything toward working on this "almost" N293 or N393 "Frankenphone", until more recently.  (Calling it a "Frankenphone" might be a little unfair on my part; it looks like the refurbishment was professionally done, to change from using the original inductor and condensers, to using the 425 type network.)

When I received the phone on September 20, it turned out that the leaf springs on the original switch were worse than what appeared in the photograph from the seller (the one included in my first post on Sept. 16).  The contact springs were irreparably bent out of shape and the main actuator spring did not even engage with the other two contact leaves.

So I recently purchased a "reproduction" four-contact hook switch assembly (from Oldphoneworks.com).  Unfortunately, this "reproduction" was not quite right: the geometry of the new switch frame was such that I could not install the original type 143 receiver hook into it.  (The pair of fulcrum pin holes in the switch frame were too far up along the sides of the frame, and too close to the hook-stop edge on the back of the frame facing the switch springs.  This prevented the hole in the hook from lining up with the holes in the new switch frame.)  Fortunately, I was able to transfer the new leaf spring pile-up from the new switch frame onto the original switch frame, and it all appears to work well, now.

The wiring analysis posted by @poplar1 on Sept. 17 was almost right.  For some unknown reason, the "refurbisher" chose to connect the switch to L1 (with the white-red wire), and added a yellow jumper wire from L1 to F (instead of just directly connecting the switch white-red wire to F).  Maybe (s)he thought that L1 was connected to something internally in the network, and had to be included in the circuit.  Also, the external RF filter "can" that @poplar1 mentioned, is not there.  The dial's pulse contacts are connected to F and RR on the 425 network, so uses the RC filter inside the network.

Another strange thing that the "refurbisher" did (or, more accurately, did not): the large "canned" condenser pair (type N2194) is left inside the phone, even though they are not connected to anything.  The two capacitors were just connected in parallel at two of the three spare terminals in the upper-right corner of the box, and nothing else was connected to them.  (The ringer and talk circuits use the respective capacitors inside the 425 network.) So why not just remove the condenser can and the extra wires connected to it?

After analysing the phone's wiring, I created a pictorial diagram (attached below; obviously not to scale, and some of the wires are routed differently than what is in the actual phone, but the connections are the same).  The diagram shows the phone's "refurbished" wiring as I received it, but with the new four-contact switch instead of the original three-contact one.  The unconnected pair of switch contacts are planned to be used to connect and power my pulse-to-tone converter circuit.

I have yet to test the phone, and to experiment with the ringer capacitor to optimise the ringing.

That's all for now.  Hopefully, I can spend a little more time on this, now.

Regards to all,
  Ed ("The DuinoSoar")

[poplar1]

You have a wire going from BK on the dial to R on the network. Rather, BK on dial should go to RR.

You also have a wire from W on the dial to RR on the network. This wire should be removed. Instead, one side of the receiver should connect to W on the dial, and the other receiver wire to R on the network. Both receiver wires should *not* connect directly to the network. The BB and W contact springs on the dial open while dialing, which disconnects the receiver so that you don't hear the loud pops in your ear while the dial is pulsing out.

[DuinoSoar]

Thank you, @poplar1, for pointing out the incorrect wiring of the dial.

The diagram in my previous post shows the way the phone is currently wired, as I received it from the seller back in September (except that it shows a four-contact switch, instead of the original three-contact one, and the large condenser is gone).  As I wrote before, I have not yet had any chance to try to connect or test the phone (which is probably a good thing since it is wired wrong).

So, whoever "refurbished" it must have thought that the pulse contacts are between the "W" and "Y" terminals (an easy mistake, I suppose, because the "W" terminal is next to one of the pulsing contact springs), and also thought that the "BK" and "BB" terminals of the off-normal shunting contacts could be used to shunt the receiver (similar to how it is done in a 5oo set).  So now I wonder if the refurbishment was done professionally, after all.

I found a diagram of the dial (or a similar one) online, and attached it below (I cleaned up the terminal labels in the picture) showing that the dial connections in my phone are completely wrong.

I also see that shunting the receiver during dial operation will not work, because the "BK" terminal is common between one of the off-normal shunt contacts (normally open) and one of the pulse contacts.  So, as you suggested, the receiver will need to be disconnected during dialling (using terminals "W" and "BB"), instead of shunted.

So, the normally-closed off-normal switch contacts between W and BB should be in series with the receiver (to disconnect the receiver during dialling) and the pulse contacts Y and BK should be connected to F and RR on the network.

I will redesign the wiring and post a new pictorial diagram of the phone; hopefully later today.

Thanks and best regards,
   Ed ("The DuinoSoar")

[DuinoSoar]

I re-wired some of the connections.  Attached below is a new pictorial drawing.

Electrical changes:

• removed the red wire from between network terminal "GN" and the receiver "box-terminal";
• moved the white wire (which is connected to dial terminal "W") from network terminal RR to the same receiver terminal mentioned in Step 1;
• moved the white-black wire (which is connected to dial terminal "BK") from network terminal "R" to network terminal "RR".

Steps 1 and 2 above connects the dial's normally-closed Off-Normal contacts between the receiver terminal and network terminal "GN" (so that the receiver is disconnected during dialling).  Step 3 above connects the dial's pulse contacts between network terminals "F" and "RR".

I also made some "cosmetic" changes (electrical connections remain the same):

• moved the yellow jumper wire from network terminal "F" to the "tip" circuit "box-terminal" in the upper-right corner of the box;
• moved the switch white-red wire from network terminal "L1" to network terminal "F";
• moved the green Tip line wire from the "tip" box-terminal to network terminal "L1";
• swapped the transmitter's black and red wires on the "door" transmitter connections, to more closely match the colours of the transmitter wires in the wiring harness.

I still need to actually solder the leads on the new switch's lug terminals before I can test the phone, but I think I am closer now.

Regards to all,
   Ed ("The Duinosoar")

[DuinoSoar]

I used my cheap multimeter (with a capacitance meter position) to measure the capacitance of the condenser between the 425B network's A, K terminals (after removing one of the red ringer wires, of course).

The capacitance reading would not settle on a singe value, and kept varying between about 400 to 600 nF.

I have a 500 set, which also has a 425 type network, so I tried measuring its A-K ringer capacitance for comparison, and it was rock-solid steady at 468 nF.  (Nominal value is either 470 nF or 500 nF, depending on which schematic is referenced.)

So I take all this to mean the that the A-K condenser in my N293's 425B network is defective.

As well, the ringer in my '293 has the two 700 Ω coils (total 1.4 KΩ), and any '239 schematic I have seen with this value of ringer coil resistance always shows a 1 µF condenser.

So my intention is to move the two red ringer-circuit wires from network terminals A, K, to the two spare "box" terminals near the upper-right corner of the box, and add two new metalised polypropylene film capacitors in parallel, between these two "box" terminals: one 470 nF and the other 560 nF (nominal values, for a total of 1.03 µF), both rated at 400 V.

QUESTION:  Will the 400 V capacitor rating be enough to prevent capacitor breakdown due to transient voltage spikes from collapsing magnetic fields in the ringer coils?  (Is this something to even worry about?)

[HarrySmith]

On the 425 network you do have blind terminals. I have used these before as attaching points in a conversion. They are L1, S, T and possibly G on older networks. Hope that helps.

[TelePlay]

QUESTION:  Will the 400 V capacitor rating be enough to prevent capacitor breakdown due to transient voltage spikes from collapsing magnetic fields in the ringer coils?  (Is this something to even worry about?)

The most often recommended value I remember seeing is 230 Volts for that capacitor. The ring AC voltage is about 90 VAC. 400 V is more than enough.

[TelePlay]

On the 425 network you do have blind terminals. I have used these before as attaching points in a conversion. They are L1, S, T and possibly G on older networks. Hope that helps.

Found this today looking for something else. L1, L2 and G are not connected to the network, A and K have the ring capacitor between them within the network, if this helps.

[DuinoSoar]

Thank you, @HarrySmith and @TelePlay.  Yes, I have schematics of the 425 network.  But the spare terminals that are mounted in the sides of the wooden box are quite close to the ringer and are handy enough for mounting the new capacitors.

And @TelePlay, my concern was not so much about the 90 Vac ring voltage, as much as voltage spikes resulting from collapsing magnetic fields in the ringer coils (often called a "flyback voltage").  Although very transient, these voltages can be much higher than the original applied voltage that built up the magnetic fields.  However, I do not really know if this should be of much concern: the transience of any flyback voltage is probably short enough to not cause any problem for the capacitors.